Investigation of Electrical Tree Growth of XLPE Nano Composites using Time–Frequency Map and Clustering Analysis of PD Signals

Abstract

Failure of underground cables due to electrical treeing phenomena, i.e. formation of electrical discharges in the imperfections of cable insulation, is a major problem faced by electrical utilities. Many research works on nanocomposites are being carried out to improve the electrical treeing resistance of the XLPE cable insulation material. Technological advancements in communication and data analytic systems have shown the way for implementing online continuous partial discharge (PD) monitoring systems for high voltage apparatus. Hence collection of PD database of XLPE nanocomposites in the laboratory during entire electrical tree growth process is important for implementing efficient condition monitoring systems and relatively little work has been published in this area. In this work, PD characteristics of XLPE nanocomposites with 1, 3, 5 and 10 wt% silica were investigated. Electrical tree growth and corresponding phase resolved PD (PRPD) pattern were analysed with respect to time. Cluster analysis of equivalent time–frequency mapping of PD signals was carried out with respect to tree growth time period. Statistical analysis was performed for the entire set of PD data. Results show that cluster analysis of T–F map of PD data is useful in estimating early failure of insulating material due to treeing. Addition of silica nano fillers in the range of 3–5 wt% concentration significantly improves the PD resistance and breakdown time of XLPE material.